SaaS economics Part 3: the missing piece
- Summary:
- Salesforce recent analyst summit provides insights to what might be the saviour of the SaaS economics business model. I am less than convinced. Let's contextualise that into Part 3 of this series.
When I penned Part 1 and Part 2 of SaaS economics, I could not have known that coincidentally, Phil Wainewright was being given an outline of how Salesforce plans to move forward. In that sense, I can't help but wonder if an alter ago was sitting on Wainewright's shoulder. Reporting on the way Salesforce sees its corner of the SaaS market going forward, Wainewright draws heavily on the investor presentation which focused on winning in the large enterprise.
It is easy to be enticed by the idea that operational scale in the form of seven and eight figure deals paves the way to continued Salesforce growth and long term profitability. And when you are counting 800 out of 150,000 companies as representing 20% of your revenue - i.e. roughly $1 billion annually - then who wouldn't go elephant hunting?
But there are flaws in Salesforce argument that need review. Let's start with Wainewright's observation that:
I’m more concerned by the implications of such a heavy focus on large enterprise deals. One of Salesforce’s strengths, it has always seemed to me, has been its appeal across-the-board to organizations of all sizes. This encourages Salesforce to ensure the growing complexity of its offerings doesn’t get in the way of maintaining ease of adoption and use.
Enterprise reality bites
Let's consider this through the enterprise lens.
When Salesforce shows up in a deal, it is not selling one cloud but a handful. Wainewright's hope that the company can offer a straight through line of sight across all these solutions doesn't stand up to scrutiny. In the few large deals I have seen, it is the very fact that Salesforce is hefting multiple clouds that may - or may not - talk to one another that gives rise to concern among IT departments well versed in understanding what this means in terms of end user support, feeder systems maintenance and outputs to other systems and analytics applications.
Make no mistake, IT is still a very important part of the decision making process, something that is not found so often in the SME space that forms the Salesforce heartland. In that context, Salesforce is very little different to any other cloud provider, large or small. Integration and the costs that go with them don't magically vanish because something is cloud. From everything I see, Salesforce is not the arms dealer in that war. Rather it is more likely to be a Dell or IBM.
In deals where the incumbent can credibly (if sometimes tenuously) talk about an integrated solution then I rate Salesforce chances of winning little better than 45% unless the business user has the final vote AND is thoroughly fed up with what they already use.
Those odds fall dramatically if you are looking at anything that has a significant manufacturing component in the enterprise applications landscape. In short, Salesforce may think it has a massive opportunity based upon Gartner metrics (see below) but you have to look very carefully into the industries that might prove attracted to its solutions and then refactor those numbers.
Note: TAM here refers to Total Addressable Market.
This puts into question the value Salesforce is able to deliver because once I as a buyer start factoring in those costs, then the Salesforce licence takes on a different complexion. Rather than being the leading cost, it may end up being one of many costs that have to be factored into a decision. One thing is for sure. Even where the deals run millions of dollars, there is no guarantee that this will provide the account control Salesforce believes will ensure long term profitability. I have seen examples where a brand leading win that looked for all intents and purposes like a rip and replace turned out to be little more than a side deal. And it will be those rip and replace deals that hold the real magic for Salesforce's future profitability.
The churn issue
The more worrying aspect of Salesforce strategy lies in its implicit acceptance of attrition rates that surprised me. According to Wainewright's report, the average runs at 10% but in enterprise runs at half that figure. That's a helluva number that Salesforce has to make up just to stand still each year.
I'd go further, based upon what I saw in the underlying analysis of numbers used to illustrate Part 1 of my SaaS economics series and confidential information I have seen from other SaaS vendors. Salesforce overall attrition rate is significantly above the industry average. Here is an example graph from that underlying research which was garnered form 306 vendors, albeit many serving SME markets:
There are many more metrics that can be included here but when you get into the kinds of term contracts Salesforce expects and at the value levels it hopes to achieve, then the overall average churn rates are looking to be much more like 6% overall. In areas that are arguably more sticky like financials, churn rates drop dramatically. Again, pulling from my own research, I typically see rates in the 3-5% range. Why is Salesforce so much higher?
I could offer a slew of reasons but Wainewright gets close to the answer where he says:
The financial metrics are clear enough — it doesn’t make sense to invest a lot of dollars in acquiring customers that may only be around for a few years. But to me that’s an argument for investing in more automated ways of bringing them on board and servicing their needs — especially in reducing those somewhat high attrition rates.
The missing piece
This plays towards an argument Naomi Bloom offers in her critique to my Part 1 piece that ended up on the LinkedIn HR Tech group. She says in relation to HR though it applies equally here:
Customer satisfaction and retention rates are driven by many factors, from having wonderful and useful product capabilities to having a very sticky user experience, and there's a lot of room here for unique approaches by different vendors and/or for different market segments. Running a brilliant operating environment means building tools for everything from provisioning to payroll scheduling, tools which cannot be bought "off the shelf" and which are themselves complex applications. So one thing I advise all buyers to consider is how far along their proposed SaaS vendor is in having industrialized every aspect of operations, for much of which you must have the right SaaS architecture in the first place.
Bloom is rigid in her architectural approach which, she argues requires: "interrogatory configuration" at its heart. This is far easier said than done and technical people will argue about its viability. Even so, Bloom believes that vendors can get there. Just getting a partial solution in place would be a good start. But this is where I believe Salesforce is weakest.
While it has done a tremendous job reinventing CRM and avoiding ghastly upgrades, Salesforce is saddled with two things: a mess of clouds it has to integrate as noted above. That plays directly counter to the Bloom/Wainewright argument. Second and despite plenty of engineering effort, Salesforce seems to be stuck with an architecture it cannot get away from without totally refactoring the solution. That would require the spending of many millions of dollars and a significant ramping of R&D at a time when the market is thirsty for profitable growth. In short, Salesforce may not be able to provide the simplicity implied by Wainewright's observations and Bloom's answer.
Contrast that with Workday, which, while having far fewer customers and in a market that is arguably more 'sticky,' has almost no churn.
Salesforce answer
As I was picking over Wainewright's report, I came across a series of slides from earlier investor presentations. There were some similarities to those shown recently and ar in line with what Wainewright saw. One in particular is interesting. It appears to position Salesforce growth as coming from analytics. See below:
We already know that Salesforce is aiming Wave directly at large enterprise. The current pricing model alone tells me that the company is excluding the bulk of its user base from this solution. I am firmly convinced that is a huge mistake. While SME customers can get plenty of reasonable analytics in Salesforce dashboards, Wave's promise is so much more for the many mobile users among Salesforce customer portfolio. If anything, users I saw at Dreamforce were wild with enthusiasm at the potential to add value to their deployments.
It would be a far easier and cost effective upsell to make Wave available across the board than by simply limiting to 800 or more customers through an exclusionary price mechanism. Depending on what numbers you choose to put into that equation, there might well be a billion dollar opportunity in enterprise today but it will still have to be sold and marketed. And it won't magically appear in the next year - or two years for that matter. An upsell at sensible rates would likely reap much larger rewards over the long term across the whole customer portfolio.
Concluding thoughts
While Salesforce is tacitly acknowledging that servicing the small business is not profitable enough, its levers appear limited with no real sign that it can overcome the associated problems. Instead, it is pinning its hopes on acquiring enterprise customers expecting those to add growth that outstrips its attrition rates among smaller business.
I hope I am wrong but on its face, that sounds like a high risk strategy that will require continued high marketing spend and against which it already has formidable opposition.
It's a gift to other players in the market like SugarCRM, Zoho and the newly re-minted Hubspot. They will see this as an opportunity to position themselves as the provider that does offer that great user experience across all CRM related applications for SMEs. From that standpoint, those companies look in better shape to meet the challenges of SaaS economics.
The real worry though comes here: if Salesforce cannot move into profitable business using SaaS economics, then who can?